Spatial and temporal characteristics of particulate matter in Beijing, China using the Empirical Mode Decomposition method

Air pollution has become a serious problem in Beijing, China. Daily PM10 mass concentration measurements were collected at 27 stations in Beijing over a 5-year period from January 1, 2008 to October 31, 2012. We used a new clustering method (kernel K-means) and a new period and trend decomposition method (Empirical Mode Decomposition, EMD) to explore the spatial and temporal characteristics of the PM10 mass concentration in the City. The temporal period and trend of each cluster center were decomposed using the EMD method, which is an adaptive data analysis method that requires no prior information. The daily PM10 mass concentrations varied greatly from 5 mu g/m(3) to more than 600 mu g/m(3). All of the stations were partitioned into three clusters by the kernel K-means method, and which represent the low-, middle- and high-pollution stations, respectively. The first cluster contained nine stations, mainly located in the north suburban area. The second cluster, whose degree of pollution was much more serious than the first cluster, contained 13 stations distributed in urban and pen-urban areas. The pollution level in the southern part of Beijing was much more serious than in the northern part of the City. The third cluster contained five stations located outside the second-cluster stations. The total decreased amplitudes of the three clusters during the whole period were 19 mu g/m(3), 10 mu g/m(3) and 4 mu g/m(3), respectively. Although the global trend of the PM10 mass concentration decreased in general, it was not the same for each season and station. The trends in summer and winter declined, while in spring, it has been increasing in recent years. Five types of trends can be found for stations, including monotonic decreasing, rise fall, fall rise fall, fall rise and rise. The rising trend of the regional background air pollution monitoring station, Miyun-reservoir, indicates an increase in the City's background PM10 mass concentration. (C) 2013 Elsevier B.V. All rights reserved.